Convolutional Neural Networks

Project: Write an Algorithm for a Dog Identification App


In this notebook, some template code has already been provided for you, and you will need to implement additional functionality to successfully complete this project. You will not need to modify the included code beyond what is requested. Sections that begin with '(IMPLEMENTATION)' in the header indicate that the following block of code will require additional functionality which you must provide. Instructions will be provided for each section, and the specifics of the implementation are marked in the code block with a 'TODO' statement. Please be sure to read the instructions carefully!

Note: Once you have completed all of the code implementations, you need to finalize your work by exporting the Jupyter Notebook as an HTML document. Before exporting the notebook to html, all of the code cells need to have been run so that reviewers can see the final implementation and output. You can then export the notebook by using the menu above and navigating to File -> Download as -> HTML (.html). Include the finished document along with this notebook as your submission.

In addition to implementing code, there will be questions that you must answer which relate to the project and your implementation. Each section where you will answer a question is preceded by a 'Question X' header. Carefully read each question and provide thorough answers in the following text boxes that begin with 'Answer:'. Your project submission will be evaluated based on your answers to each of the questions and the implementation you provide.

Note: Code and Markdown cells can be executed using the Shift + Enter keyboard shortcut. Markdown cells can be edited by double-clicking the cell to enter edit mode.

The rubric contains optional "Stand Out Suggestions" for enhancing the project beyond the minimum requirements. If you decide to pursue the "Stand Out Suggestions", you should include the code in this Jupyter notebook.


Why We're Here

In this notebook, you will make the first steps towards developing an algorithm that could be used as part of a mobile or web app. At the end of this project, your code will accept any user-supplied image as input. If a dog is detected in the image, it will provide an estimate of the dog's breed. If a human is detected, it will provide an estimate of the dog breed that is most resembling. The image below displays potential sample output of your finished project (... but we expect that each student's algorithm will behave differently!).

Sample Dog Output

In this real-world setting, you will need to piece together a series of models to perform different tasks; for instance, the algorithm that detects humans in an image will be different from the CNN that infers dog breed. There are many points of possible failure, and no perfect algorithm exists. Your imperfect solution will nonetheless create a fun user experience!

The Road Ahead

We break the notebook into separate steps. Feel free to use the links below to navigate the notebook.

  • Step 0: Import Datasets
  • Step 1: Detect Humans
  • Step 2: Detect Dogs
  • Step 3: Create a CNN to Classify Dog Breeds (from Scratch)
  • Step 4: Create a CNN to Classify Dog Breeds (using Transfer Learning)
  • Step 5: Write your Algorithm
  • Step 6: Test Your Algorithm

Step 0: Import Datasets

Make sure that you've downloaded the required human and dog datasets:

Note: if you are using the Udacity workspace, you DO NOT need to re-download these - they can be found in the /data folder as noted in the cell below.

  • Download the dog dataset. Unzip the folder and place it in this project's home directory, at the location /dog_images.

  • Download the human dataset. Unzip the folder and place it in the home directory, at location /lfw.

Note: If you are using a Windows machine, you are encouraged to use 7zip to extract the folder.

In the code cell below, we save the file paths for both the human (LFW) dataset and dog dataset in the numpy arrays human_files and dog_files.

In [1]:
import numpy as np
from glob import glob
PATH = "C:\\Users\\krishna\\Desktop\\Data Science\\DeepLearning\\Udacity\\"

# load filenames for human and dog images
human_files = np.array(glob("C:\\Users\\krishna\\Desktop\\Data Science\\DeepLearning\\Udacity\\lfw\\*\\*"))
dog_files = np.array(glob("C:\\Users\\krishna\\Desktop\\Data Science\\DeepLearning\\Udacity\\dogImages\\*\\*\\*"))

# print number of images in each dataset
print('There are %d total human images.' % len(human_files))
print('There are %d total dog images.' % len(dog_files))
There are 13233 total human images.
There are 8351 total dog images.

Step 1: Detect Humans

In this section, we use OpenCV's implementation of Haar feature-based cascade classifiers to detect human faces in images.

OpenCV provides many pre-trained face detectors, stored as XML files on github. We have downloaded one of these detectors and stored it in the haarcascades directory. In the next code cell, we demonstrate how to use this detector to find human faces in a sample image.

In [2]:
import cv2                
import matplotlib.pyplot as plt                        
%matplotlib inline                               

#local path: C:/Users/krishna/Desktop/Data Science/DeepLearning/Udacity/opencv/data/
# extract pre-trained face detector
face_cascade = cv2.CascadeClassifier('haarcascade_frontalface_alt.xml')

# load color (BGR) image
img = cv2.imread(human_files[0])
# convert BGR image to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# find faces in image
faces = face_cascade.detectMultiScale(gray)

# print number of faces detected in the image
print('Number of faces detected:', len(faces))

# get bounding box for each detected face
for (x,y,w,h) in faces:
    # add bounding box to color image
    cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
    
# convert BGR image to RGB for plotting
cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

# display the image, along with bounding box
plt.imshow(cv_rgb)
plt.show()
Number of faces detected: 1

Before using any of the face detectors, it is standard procedure to convert the images to grayscale. The detectMultiScale function executes the classifier stored in face_cascade and takes the grayscale image as a parameter.

In the above code, faces is a numpy array of detected faces, where each row corresponds to a detected face. Each detected face is a 1D array with four entries that specifies the bounding box of the detected face. The first two entries in the array (extracted in the above code as x and y) specify the horizontal and vertical positions of the top left corner of the bounding box. The last two entries in the array (extracted here as w and h) specify the width and height of the box.

Write a Human Face Detector

We can use this procedure to write a function that returns True if a human face is detected in an image and False otherwise. This function, aptly named face_detector, takes a string-valued file path to an image as input and appears in the code block below.

In [3]:
# returns "True" if face is detected in image stored at img_path
def face_detector(img_path):
    img = cv2.imread(img_path)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    faces = face_cascade.detectMultiScale(gray)
    return len(faces) > 0

(IMPLEMENTATION) Assess the Human Face Detector

Question 1: Use the code cell below to test the performance of the face_detector function.

  • What percentage of the first 100 images in human_files have a detected human face?
  • What percentage of the first 100 images in dog_files have a detected human face?

Ideally, we would like 100% of human images with a detected face and 0% of dog images with a detected face. You will see that our algorithm falls short of this goal, but still gives acceptable performance. We extract the file paths for the first 100 images from each of the datasets and store them in the numpy arrays human_files_short and dog_files_short.

Answer: (You can print out your results and/or write your percentages in this cell)

96 % of human_files have a detected human face
18 % of dog_files have a detected human face

In [4]:
from tqdm import tqdm

human_files_short = human_files[:100]
dog_files_short = dog_files[:100]

#-#-# Do NOT modify the code above this line. #-#-#

## TODO: Test the performance of the face_detector algorithm 
## on the images in human_files_short and dog_files_short.
human_count = 0
dog_count = 0
for i in range(0,100):
    if face_detector(human_files_short[i]):
        human_count+=1
    if face_detector(dog_files_short[i]):
        dog_count+=1
print(human_count,"% of human_files have a detected human face")
print(dog_count,"% of dog_files have a detected human face")
96 % of human_files have a detected human face
18 % of dog_files have a detected human face

We suggest the face detector from OpenCV as a potential way to detect human images in your algorithm, but you are free to explore other approaches, especially approaches that make use of deep learning :). Please use the code cell below to design and test your own face detection algorithm. If you decide to pursue this optional task, report performance on human_files_short and dog_files_short.

In [5]:
### (Optional) 
### TODO: Test performance of anotherface detection algorithm.
### Feel free to use as many code cells as needed.

Step 2: Detect Dogs

In this section, we use a pre-trained model to detect dogs in images.

Obtain Pre-trained VGG-16 Model

The code cell below downloads the VGG-16 model, along with weights that have been trained on ImageNet, a very large, very popular dataset used for image classification and other vision tasks. ImageNet contains over 10 million URLs, each linking to an image containing an object from one of 1000 categories.

In [57]:
import torch
import torchvision.models as models

# define VGG16 model
VGG16 = models.vgg16(pretrained=True)

# check if CUDA is available
use_cuda = torch.cuda.is_available()

# move model to GPU if CUDA is available
if use_cuda:
    VGG16 = VGG16.cuda()

Given an image, this pre-trained VGG-16 model returns a prediction (derived from the 1000 possible categories in ImageNet) for the object that is contained in the image.

(IMPLEMENTATION) Making Predictions with a Pre-trained Model

In the next code cell, you will write a function that accepts a path to an image (such as 'dogImages/train/001.Affenpinscher/Affenpinscher_00001.jpg') as input and returns the index corresponding to the ImageNet class that is predicted by the pre-trained VGG-16 model. The output should always be an integer between 0 and 999, inclusive.

Before writing the function, make sure that you take the time to learn how to appropriately pre-process tensors for pre-trained models in the PyTorch documentation.

In [7]:
print(VGG16)
VGG(
  (features): Sequential(
    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (1): ReLU(inplace=True)
    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (3): ReLU(inplace=True)
    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (6): ReLU(inplace=True)
    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (8): ReLU(inplace=True)
    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (11): ReLU(inplace=True)
    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (13): ReLU(inplace=True)
    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (15): ReLU(inplace=True)
    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (18): ReLU(inplace=True)
    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (20): ReLU(inplace=True)
    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (22): ReLU(inplace=True)
    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (25): ReLU(inplace=True)
    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (27): ReLU(inplace=True)
    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (29): ReLU(inplace=True)
    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
  )
  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))
  (classifier): Sequential(
    (0): Linear(in_features=25088, out_features=4096, bias=True)
    (1): ReLU(inplace=True)
    (2): Dropout(p=0.5, inplace=False)
    (3): Linear(in_features=4096, out_features=4096, bias=True)
    (4): ReLU(inplace=True)
    (5): Dropout(p=0.5, inplace=False)
    (6): Linear(in_features=4096, out_features=1000, bias=True)
  )
)
In [61]:
from PIL import Image
import torchvision.transforms as transforms

def VGG16_predict(img_path):
    '''
    Use pre-trained VGG-16 model to obtain index corresponding to 
    predicted ImageNet class for image at specified path
    
    Args:
        img_path: path to an image
        
    Returns:
        Index corresponding to VGG-16 model's prediction
    ''' 
    
    ## TODO: Complete the function.
    ## Load and pre-process an image from the given img_path
    ## Return the *index* of the predicted class for that image
    #resize the image to 224x224 as this is the shape that VGG accepts.  we can also use RandomResizedCrop(224)
    img_transform = transforms.Compose([transforms.Resize(size = (224,224)), 
                                      transforms.ToTensor()])

    image = Image.open(img_path)
    image_transformed = img_transform(image).unsqueeze(0)
    
    #pul VGG in eval mode
    VGG16.eval()

    if use_cuda:
        image_transformed = image_transformed.cuda()
    
    output = VGG16(image_transformed)
    _, preds_tensor = torch.max(output,1)
    preds = np.squeeze(preds_tensor.numpy()) if not use_cuda else np.squeeze(preds_tensor.cpu().numpy())
    
    return preds # predicted class index
In [62]:
VGG16_predict(dog_files_short[0])
Out[62]:
array(552, dtype=int64)

(IMPLEMENTATION) Write a Dog Detector

While looking at the dictionary, you will notice that the categories corresponding to dogs appear in an uninterrupted sequence and correspond to dictionary keys 151-268, inclusive, to include all categories from 'Chihuahua' to 'Mexican hairless'. Thus, in order to check to see if an image is predicted to contain a dog by the pre-trained VGG-16 model, we need only check if the pre-trained model predicts an index between 151 and 268 (inclusive).

Use these ideas to complete the dog_detector function below, which returns True if a dog is detected in an image (and False if not).

In [10]:
### returns "True" if a dog is detected in the image stored at img_path
def dog_detector(img_path):
    ## TODO: Complete the function.
    preds = VGG16_predict(img_path)
    
    return (preds >=151 and preds <= 268)

(IMPLEMENTATION) Assess the Dog Detector

Question 2: Use the code cell below to test the performance of your dog_detector function.

  • What percentage of the images in human_files_short have a detected dog?
  • What percentage of the images in dog_files_short have a detected dog?

Answer:

0% of human_files_short have a detected dog face
93% of dog_files_short have a detected dog face

In [88]:
### TODO: Test the performance of the dog_detector function
### on the images in human_files_short and dog_files_short.
from tqdm import tqdm

human_files_short = human_files[:100]
dog_files_short = dog_files[:100]

#-#-# Do NOT modify the code above this line. #-#-#

## TODO: Test the performance of the face_detector algorithm 
## on the images in human_files_short and dog_files_short.
human_count = 0
dog_count = 0
for i in range(0,100):
    if dog_detector(human_files_short[i]):
        human_count+=1
    if dog_detector(dog_files_short[i]):
        dog_count+=1
print(human_count,"% of human_files have a detected dog face")
print(dog_count,"% of dog_files have a detected dog face")
0 % of human_files have a detected dog face
93 % of dog_files have a detected dog face

We suggest VGG-16 as a potential network to detect dog images in your algorithm, but you are free to explore other pre-trained networks (such as Inception-v3, ResNet-50, etc). Please use the code cell below to test other pre-trained PyTorch models. If you decide to pursue this optional task, report performance on human_files_short and dog_files_short.

In [12]:
### (Optional) 
### TODO: Report the performance of another pre-trained network.
### Feel free to use as many code cells as needed.

Step 3: Create a CNN to Classify Dog Breeds (from Scratch)

Now that we have functions for detecting humans and dogs in images, we need a way to predict breed from images. In this step, you will create a CNN that classifies dog breeds. You must create your CNN from scratch (so, you can't use transfer learning yet!), and you must attain a test accuracy of at least 10%. In Step 4 of this notebook, you will have the opportunity to use transfer learning to create a CNN that attains greatly improved accuracy.

We mention that the task of assigning breed to dogs from images is considered exceptionally challenging. To see why, consider that even a human would have trouble distinguishing between a Brittany and a Welsh Springer Spaniel.

Brittany Welsh Springer Spaniel

It is not difficult to find other dog breed pairs with minimal inter-class variation (for instance, Curly-Coated Retrievers and American Water Spaniels).

Curly-Coated Retriever American Water Spaniel

Likewise, recall that labradors come in yellow, chocolate, and black. Your vision-based algorithm will have to conquer this high intra-class variation to determine how to classify all of these different shades as the same breed.

Yellow Labrador Chocolate Labrador Black Labrador

We also mention that random chance presents an exceptionally low bar: setting aside the fact that the classes are slightly imabalanced, a random guess will provide a correct answer roughly 1 in 133 times, which corresponds to an accuracy of less than 1%.

Remember that the practice is far ahead of the theory in deep learning. Experiment with many different architectures, and trust your intuition. And, of course, have fun!

(IMPLEMENTATION) Specify Data Loaders for the Dog Dataset

Use the code cell below to write three separate data loaders for the training, validation, and test datasets of dog images (located at dog_images/train, dog_images/valid, and dog_images/test, respectively). You may find this documentation on custom datasets to be a useful resource. If you are interested in augmenting your training and/or validation data, check out the wide variety of transforms!

In [13]:
from glob import glob

dog_files = np.array(glob("C:\\Users\\krishna\\Desktop\\Data Science\\DeepLearning\\Udacity\\dogImages\\*\\*\\*"))
In [89]:
import os
from torchvision import datasets


### TODO: Write data loaders for training, validation, and test sets
## Specify appropriate transforms, and batch_sizes

data_dir = 'C:/Users/krishna/Desktop/Data Science/DeepLearning/Udacity/dogImages/'

train_dir = os.path.join(data_dir, 'train/')
val_dir = os.path.join(data_dir, 'valid/')
test_dir = os.path.join(data_dir, 'test/')

data_transform_train = transforms.Compose([transforms.RandomResizedCrop(224), 
                                     transforms.RandomHorizontalFlip(), 
                                     transforms.RandomRotation(15), 
                                     transforms.ToTensor(),
                                     transforms.Normalize((0.475, 0.445, 0.408), (0.224, 0.212, 0.219))])

data_transform = transforms.Compose([transforms.RandomResizedCrop(224),  
                                     transforms.ToTensor(),
                                     transforms.Normalize((0.475, 0.445, 0.408), (0.224, 0.212, 0.219))])

train_data = datasets.ImageFolder(train_dir, transform=data_transform_train)
val_data = datasets.ImageFolder(val_dir, transform=data_transform)
test_data = datasets.ImageFolder(test_dir, transform=data_transform)

print('Number training images: ', len(train_data))
print('Number validation images: ', len(val_data))
print('Number test images: ', len(test_data))
Number training images:  6680
Number validation images:  835
Number test images:  836
In [15]:
# number of subprocesses to use for data loading
num_workers = 0
# how many samples per batch to load
batch_size = 20

# prepare data loaders (combine dataset and sampler)
train_loader = torch.utils.data.DataLoader(train_data, batch_size=batch_size, num_workers=num_workers, shuffle=True)
valid_loader = torch.utils.data.DataLoader(val_data, batch_size=batch_size, num_workers=num_workers)
test_loader = torch.utils.data.DataLoader(test_data, batch_size=batch_size, num_workers=num_workers)

loaders_scratch = {'train' : train_loader, 'valid' : valid_loader, 'test' : test_loader}

Question 3: Describe your chosen procedure for preprocessing the data.

  • How does your code resize the images (by cropping, stretching, etc)? What size did you pick for the input tensor, and why?
  • Did you decide to augment the dataset? If so, how (through translations, flips, rotations, etc)? If not, why not?

Answer:

The resize was handled with RandomResizedCrop() with 224 as cropped image size as the VGG16 accepts input image size of 224.
I have decided to augment the data through HorizantalFlip and random rotation of 15degrees.

(IMPLEMENTATION) Model Architecture

Create a CNN to classify dog breed. Use the template in the code cell below.

In [16]:
import torch.nn as nn
import torch.nn.functional as F

# define the CNN architecture
class Net(nn.Module):
    ### TODO: choose an architecture, and complete the class
    def __init__(self):
        super(Net, self).__init__()
        ## Define layers of a CNN
        # convolutional layer (sees 224x224x3 image tensor)
        self.conv1 = nn.Conv2d(3, 32, 3, stride=2, padding=1)
        # convolutional layer (sees 56x56x32 image tensor)
        self.conv2 = nn.Conv2d(32, 64, 3, stride=2, padding=1)
        # convolutional layer (sees 14x14x64 image tensor)
        self.conv3 = nn.Conv2d(64, 128, 3, padding=1)
        
        #max pooling layer
        self.pool = nn.MaxPool2d(2,2)
        #linear layer (128*7*7 -> 500)
        self.fc1 = nn.Linear(128*7*7, 500)
        #linear layer (500->133)
        self.fc2 = nn.Linear(500, 133)
        #dropout probability of 25%
        self.dropout = nn.Dropout(0.25)
                
    
    def forward(self, x):
        ## Define forward behavior
        x = self.pool(F.relu(self.conv1(x)))
        x = self.pool(F.relu(self.conv2(x)))
        x = self.pool(F.relu(self.conv3(x)))
       
        
        x = x.view(-1, 128*7*7)
        x = self.dropout(x)
        
        x = F.relu(self.fc1(x))
        x = self.dropout(x)
        x = self.fc2(x)
        
        return x

#-#-# You so NOT have to modify the code below this line. #-#-#

# instantiate the CNN
model_scratch = Net()

# move tensors to GPU if CUDA is available
if use_cuda:
    model_scratch.cuda()

print(model_scratch)
Net(
  (conv1): Conv2d(3, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
  (conv2): Conv2d(32, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1))
  (conv3): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
  (fc1): Linear(in_features=6272, out_features=500, bias=True)
  (fc2): Linear(in_features=500, out_features=133, bias=True)
  (dropout): Dropout(p=0.25, inplace=False)
)

Question 4: Outline the steps you took to get to your final CNN architecture and your reasoning at each step.

Answer:

The image start with a 224x224x3 size.
after the first conv layer, which increases the depth to 32 with a stride of 2, the image size will be changed to 112x112x32
which then will be passed through maxPool and size becomes 56x56x32
the second conv layer will increases the depth to 64 and with a stride 2, the image size becomes 28x28x64
again with maxpool, the size will be reduced to 14x14x64
the final conv layer then increases the depth to 128 and since we going with a stride one, the size becomes 14x14x128
with one more maxpool layer, the final size becomes 7x7x128 before it will be flatten out and fed to the fully connected layer.

the fully connected layer takes the 7x7x128 image and converts to 500 features which will be fed to the output layer which outputs 133 class outputs.


I have used a dropout probability of .25 for avoiding overfitting.


Note: This is the architecture I have decided after few iterations of trial and error. I have also tried 5 conv layers with default 1 stride and that increased the training time but didnot fare better.

(IMPLEMENTATION) Specify Loss Function and Optimizer

Use the next code cell to specify a loss function and optimizer. Save the chosen loss function as criterion_scratch, and the optimizer as optimizer_scratch below.

In [17]:
import torch.optim as optim

### TODO: select loss function
criterion_scratch = nn.CrossEntropyLoss()

### TODO: select optimizer
optimizer_scratch = optim.SGD(model_scratch.parameters(), lr = 0.05)

(IMPLEMENTATION) Train and Validate the Model

Train and validate your model in the code cell below. Save the final model parameters at filepath 'model_scratch.pt'.

In [18]:
from PIL import ImageFile
ImageFile.LOAD_TRUNCATED_IMAGES = True

def train(n_epochs, loaders, model, optimizer, criterion, use_cuda, save_path):
    """returns trained model"""
    # initialize tracker for minimum validation loss
    valid_loss_min = np.Inf 
    
    if use_cuda:
        print("Training on GPU")
    else:
        print("check your GPU")
    
    for epoch in range(1, n_epochs+1):
        # initialize variables to monitor training and validation loss
        train_loss = 0.0
        valid_loss = 0.0
        
        ###################
        # train the model #
        ###################
        model.train()
        for batch_idx, (data, target) in enumerate(loaders['train']):
            # move to GPU
            if use_cuda:
                data, target = data.cuda(), target.cuda()
            ## find the loss and update the model parameters accordingly
            optimizer.zero_grad()
            output = model(data)
            
            loss = criterion(output, target)    
            loss.backward()
            optimizer.step()
            
            ## record the average training loss            
            train_loss = train_loss + ((1 / (batch_idx + 1)) * (loss.data - train_loss))
            
        ######################    
        # validate the model #
        ######################
        model.eval()
        for batch_idx, (data, target) in enumerate(loaders['valid']):
            # move to GPU
            if use_cuda:
                data, target = data.cuda(), target.cuda()
            ## update the average validation loss
            val_output = model(data)
            #calculate the batch loss
            val_loss = criterion(val_output, target)
            
            #update average validation loss
            valid_loss += ((1 / (batch_idx + 1)) * (val_loss.data - valid_loss))
        
        # print training/validation statistics 
        print('Epoch: {} \tTraining Loss: {:.6f} \tValidation Loss: {:.6f}'.format(epoch, train_loss, valid_loss))
    
        ## TODO: save the model if validation loss has decreased
        if valid_loss <= valid_loss_min:
            print('Validation loss decreased ({:.6f} --> {:.6f}).  Saving model ...'.format(
            valid_loss_min,
            valid_loss))
            torch.save(model.state_dict(), 'model_scratch.pt')
            valid_loss_min = valid_loss
    # return trained model
    return model
In [20]:
# train the model
model_scratch = train(20, loaders_scratch, model_scratch, optimizer_scratch, 
                      criterion_scratch, use_cuda, 'model_scratch.pt')
Training on GPU
Epoch: 1 	Training Loss: 4.405053 	Validation Loss: 4.383708
Validation loss decreased (inf --> 4.383708).  Saving model ...
Epoch: 2 	Training Loss: 4.342687 	Validation Loss: 4.342330
Validation loss decreased (4.383708 --> 4.342330).  Saving model ...
Epoch: 3 	Training Loss: 4.309982 	Validation Loss: 4.315364
Validation loss decreased (4.342330 --> 4.315364).  Saving model ...
Epoch: 4 	Training Loss: 4.250631 	Validation Loss: 4.284379
Validation loss decreased (4.315364 --> 4.284379).  Saving model ...
Epoch: 5 	Training Loss: 4.197515 	Validation Loss: 4.197108
Validation loss decreased (4.284379 --> 4.197108).  Saving model ...
Epoch: 6 	Training Loss: 4.159961 	Validation Loss: 4.262145
Epoch: 7 	Training Loss: 4.113922 	Validation Loss: 4.147303
Validation loss decreased (4.197108 --> 4.147303).  Saving model ...
Epoch: 8 	Training Loss: 4.056258 	Validation Loss: 4.130092
Validation loss decreased (4.147303 --> 4.130092).  Saving model ...
Epoch: 9 	Training Loss: 4.002645 	Validation Loss: 4.138555
Epoch: 10 	Training Loss: 3.973702 	Validation Loss: 4.083739
Validation loss decreased (4.130092 --> 4.083739).  Saving model ...
Epoch: 11 	Training Loss: 3.930516 	Validation Loss: 4.226418
Epoch: 12 	Training Loss: 3.891015 	Validation Loss: 3.983974
Validation loss decreased (4.083739 --> 3.983974).  Saving model ...
Epoch: 13 	Training Loss: 3.843189 	Validation Loss: 3.953241
Validation loss decreased (3.983974 --> 3.953241).  Saving model ...
Epoch: 14 	Training Loss: 3.819777 	Validation Loss: 4.118236
Epoch: 15 	Training Loss: 3.744352 	Validation Loss: 3.925568
Validation loss decreased (3.953241 --> 3.925568).  Saving model ...
Epoch: 16 	Training Loss: 3.750088 	Validation Loss: 3.982347
Epoch: 17 	Training Loss: 3.732667 	Validation Loss: 3.832092
Validation loss decreased (3.925568 --> 3.832092).  Saving model ...
Epoch: 18 	Training Loss: 3.685048 	Validation Loss: 3.954244
Epoch: 19 	Training Loss: 3.620196 	Validation Loss: 3.921146
Epoch: 20 	Training Loss: 3.592892 	Validation Loss: 3.855658
In [21]:
# load the model that got the best validation accuracy
model_scratch.load_state_dict(torch.load('model_scratch.pt'))
Out[21]:
<All keys matched successfully>

(IMPLEMENTATION) Test the Model

Try out your model on the test dataset of dog images. Use the code cell below to calculate and print the test loss and accuracy. Ensure that your test accuracy is greater than 10%.

In [22]:
def test(loaders, model, criterion, use_cuda):

    # monitor test loss and accuracy
    test_loss = 0.
    correct = 0.
    total = 0.

    model.eval()
    for batch_idx, (data, target) in enumerate(loaders['test']):
        # move to GPU
        if use_cuda:
            data, target = data.cuda(), target.cuda()
        # forward pass: compute predicted outputs by passing inputs to the model
        output = model(data)
        # calculate the loss
        loss = criterion(output, target)
        # update average test loss 
        test_loss = test_loss + ((1 / (batch_idx + 1)) * (loss.data - test_loss))
        # convert output probabilities to predicted class
        pred = output.data.max(1, keepdim=True)[1]
        # compare predictions to true label
        correct += np.sum(np.squeeze(pred.eq(target.data.view_as(pred))).cpu().numpy())
        total += data.size(0)
            
    print('Test Loss: {:.6f}\n'.format(test_loss))

    print('\nTest Accuracy: %2d%% (%2d/%2d)' % (
        100. * correct / total, correct, total))

# call test function    
test(loaders_scratch, model_scratch, criterion_scratch, use_cuda)
Test Loss: 3.972589


Test Accuracy: 11% (92/836)

Step 4: Create a CNN to Classify Dog Breeds (using Transfer Learning)

You will now use transfer learning to create a CNN that can identify dog breed from images. Your CNN must attain at least 60% accuracy on the test set.

(IMPLEMENTATION) Specify Data Loaders for the Dog Dataset

Use the code cell below to write three separate data loaders for the training, validation, and test datasets of dog images (located at dogImages/train, dogImages/valid, and dogImages/test, respectively).

If you like, you are welcome to use the same data loaders from the previous step, when you created a CNN from scratch.

In [23]:
## TODO: Specify data loaders
loaders_transfer = loaders_scratch.copy()

(IMPLEMENTATION) Model Architecture

Use transfer learning to create a CNN to classify dog breed. Use the code cell below, and save your initialized model as the variable model_transfer.

In [24]:
import torchvision.models as models
import torch.nn as nn



## TODO: Specify model architecture 
model_transfer = models.vgg16(pretrained=True)

# Freeze training for all "features" layers
for param in model_transfer.features.parameters():
    param.requires_grad = False

n_inputs = model_transfer.classifier[6].in_features
last_layer = nn.Linear(n_inputs, 133)

model_transfer.classifier[6] = last_layer


if use_cuda:
    model_transfer = model_transfer.cuda()

Question 5: Outline the steps you took to get to your final CNN architecture and your reasoning at each step. Describe why you think the architecture is suitable for the current problem.

Answer:

I picked a pre-trained VGG16 as a transfer model because it mostly provides better results. I could have tried ResNet too which also performed better on Image Classification. I checked the documentation of VGG16 to understand the accepted parameters and the input/output sizes.

I've taken the final classifer[6] and changed the output classes to 133 which is the number of classes that we have in dog breed dataset

(IMPLEMENTATION) Specify Loss Function and Optimizer

Use the next code cell to specify a loss function and optimizer. Save the chosen loss function as criterion_transfer, and the optimizer as optimizer_transfer below.

In [25]:
criterion_transfer = nn.CrossEntropyLoss()
optimizer_transfer = optim.SGD(model_transfer.parameters(), lr=0.001)

(IMPLEMENTATION) Train and Validate the Model

Train and validate your model in the code cell below. Save the final model parameters at filepath 'model_transfer.pt'.

In [26]:
from PIL import ImageFile
ImageFile.LOAD_TRUNCATED_IMAGES = True

def transfer_train(n_epochs, loaders, model, optimizer, criterion, use_cuda, save_path):
    """returns trained model"""
    # initialize tracker for minimum validation loss
    valid_loss_min = np.Inf 
    
    if use_cuda:
        print("Training on GPU")
    else:
        print("check your GPU")
    
    for epoch in range(1, n_epochs+1):
        # initialize variables to monitor training and validation loss
        train_loss = 0.0
        valid_loss = 0.0
        
        ###################
        # train the model #
        ###################
        model.train()
        for batch_idx, (data, target) in enumerate(loaders['train']):
            # move to GPU
            if use_cuda:
                data, target = data.cuda(), target.cuda()
            ## find the loss and update the model parameters accordingly
            optimizer.zero_grad()
            output = model(data)
            
            loss = criterion(output, target)    
            loss.backward()
            optimizer.step()
            
            ## record the average training loss            
            train_loss = train_loss + ((1 / (batch_idx + 1)) * (loss.data - train_loss))
            
        ######################    
        # validate the model #
        ######################
        model.eval()
        for batch_idx, (data, target) in enumerate(loaders['valid']):
            # move to GPU
            if use_cuda:
                data, target = data.cuda(), target.cuda()
            ## update the average validation loss
            val_output = model(data)
            #calculate the batch loss
            val_loss = criterion(val_output, target)
            
            #update average validation loss
            valid_loss += ((1 / (batch_idx + 1)) * (val_loss.data - valid_loss))
        
        # print training/validation statistics 
        print('Epoch: {} \tTraining Loss: {:.6f} \tValidation Loss: {:.6f}'.format(epoch, train_loss, valid_loss))
    
        ## TODO: save the model if validation loss has decreased
        if valid_loss <= valid_loss_min:
            print('Validation loss decreased ({:.6f} --> {:.6f}).  Saving model ...'.format(
            valid_loss_min,
            valid_loss))
            torch.save(model.state_dict(), 'model_transfer.pt')
            valid_loss_min = valid_loss
    # return trained model
    return model
In [27]:
# train the model
transfer_train(20, loaders_transfer, model_transfer, optimizer_transfer, criterion_transfer, use_cuda, 'model_transfer.pt')
Training on GPU
Epoch: 1 	Training Loss: 4.477378 	Validation Loss: 3.680382
Validation loss decreased (inf --> 3.680382).  Saving model ...
Epoch: 2 	Training Loss: 3.222808 	Validation Loss: 2.271841
Validation loss decreased (3.680382 --> 2.271841).  Saving model ...
Epoch: 3 	Training Loss: 2.316453 	Validation Loss: 1.643238
Validation loss decreased (2.271841 --> 1.643238).  Saving model ...
Epoch: 4 	Training Loss: 1.919255 	Validation Loss: 1.346923
Validation loss decreased (1.643238 --> 1.346923).  Saving model ...
Epoch: 5 	Training Loss: 1.718156 	Validation Loss: 1.284173
Validation loss decreased (1.346923 --> 1.284173).  Saving model ...
Epoch: 6 	Training Loss: 1.595790 	Validation Loss: 1.158978
Validation loss decreased (1.284173 --> 1.158978).  Saving model ...
Epoch: 7 	Training Loss: 1.521501 	Validation Loss: 1.190191
Epoch: 8 	Training Loss: 1.462772 	Validation Loss: 1.006248
Validation loss decreased (1.158978 --> 1.006248).  Saving model ...
Epoch: 9 	Training Loss: 1.414609 	Validation Loss: 1.114100
Epoch: 10 	Training Loss: 1.368070 	Validation Loss: 1.028024
Epoch: 11 	Training Loss: 1.302121 	Validation Loss: 1.024886
Epoch: 12 	Training Loss: 1.283345 	Validation Loss: 0.997944
Validation loss decreased (1.006248 --> 0.997944).  Saving model ...
Epoch: 13 	Training Loss: 1.281787 	Validation Loss: 1.020663
Epoch: 14 	Training Loss: 1.238328 	Validation Loss: 0.976053
Validation loss decreased (0.997944 --> 0.976053).  Saving model ...
Epoch: 15 	Training Loss: 1.239270 	Validation Loss: 0.996717
Epoch: 16 	Training Loss: 1.230007 	Validation Loss: 0.971275
Validation loss decreased (0.976053 --> 0.971275).  Saving model ...
Epoch: 17 	Training Loss: 1.193209 	Validation Loss: 0.946881
Validation loss decreased (0.971275 --> 0.946881).  Saving model ...
Epoch: 18 	Training Loss: 1.191824 	Validation Loss: 0.997940
Epoch: 19 	Training Loss: 1.183768 	Validation Loss: 0.941161
Validation loss decreased (0.946881 --> 0.941161).  Saving model ...
Epoch: 20 	Training Loss: 1.152243 	Validation Loss: 0.983634
Out[27]:
VGG(
  (features): Sequential(
    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (1): ReLU(inplace=True)
    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (3): ReLU(inplace=True)
    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (6): ReLU(inplace=True)
    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (8): ReLU(inplace=True)
    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (11): ReLU(inplace=True)
    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (13): ReLU(inplace=True)
    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (15): ReLU(inplace=True)
    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (18): ReLU(inplace=True)
    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (20): ReLU(inplace=True)
    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (22): ReLU(inplace=True)
    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (25): ReLU(inplace=True)
    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (27): ReLU(inplace=True)
    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
    (29): ReLU(inplace=True)
    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
  )
  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))
  (classifier): Sequential(
    (0): Linear(in_features=25088, out_features=4096, bias=True)
    (1): ReLU(inplace=True)
    (2): Dropout(p=0.5, inplace=False)
    (3): Linear(in_features=4096, out_features=4096, bias=True)
    (4): ReLU(inplace=True)
    (5): Dropout(p=0.5, inplace=False)
    (6): Linear(in_features=4096, out_features=133, bias=True)
  )
)
In [28]:
# load the model that got the best validation accuracy (uncomment the line below)
model_transfer.load_state_dict(torch.load('model_transfer.pt'))
Out[28]:
<All keys matched successfully>

(IMPLEMENTATION) Test the Model

Try out your model on the test dataset of dog images. Use the code cell below to calculate and print the test loss and accuracy. Ensure that your test accuracy is greater than 60%.

In [30]:
test(loaders_transfer, model_transfer, criterion_transfer, use_cuda)
Test Loss: 0.992100


Test Accuracy: 70% (590/836)

(IMPLEMENTATION) Predict Dog Breed with the Model

Write a function that takes an image path as input and returns the dog breed (Affenpinscher, Afghan hound, etc) that is predicted by your model.

In [43]:
### TODO: Write a function that takes a path to an image as input
### and returns the dog breed that is predicted by the model.

# list of class names by index, i.e. a name can be accessed like class_names[0]
class_names = [item[4:].replace("_", " ") for item in loaders_transfer['train'].dataset.classes]

def load_image(img_path):
    image = Image.open(img_path).convert('RGB')
    prediction_transform = transforms.Compose([transforms.Resize(size=(224, 224)),
                                     transforms.ToTensor(), 
                                     transforms.Normalize((0.475, 0.445, 0.408), (0.224, 0.212, 0.219))])

    # discard the transparent, alpha channel (that's the :3) and add the batch dimension
    image = prediction_transform(image)[:3,:,:].unsqueeze(0)
    return image

def predict_breed_transfer(model, class_names, img_path):
    # load the image and return the predicted breed
    img = load_image(img_path)
    model = model.cpu()
    model.eval()
    idx = torch.argmax(model(img))
    return class_names[idx]
In [92]:
new_path = "C:/Users/krishna/Desktop/Data Science/DeepLearning/Udacity/dogImages/new/"

for img_file in os.listdir(new_path):
    img_path = os.path.join(new_path, img_file)
    preds = predict_breed_transfer(model_transfer, class_names, img_path)
    
    img = Image.open(img_path)
    plt.imshow(img)
    plt.show()
    print("Predition Breed: ",preds)
Predition Breed:  American eskimo dog
Predition Breed:  Airedale terrier
Predition Breed:  Airedale terrier
Predition Breed:  American water spaniel
Predition Breed:  Australian shepherd
Predition Breed:  Basenji
Predition Breed:  Cane corso
Predition Breed:  Cavalier king charles spaniel
Predition Breed:  German wirehaired pointer
Predition Breed:  Greyhound

Step 5: Write your Algorithm

Write an algorithm that accepts a file path to an image and first determines whether the image contains a human, dog, or neither. Then,

  • if a dog is detected in the image, return the predicted breed.
  • if a human is detected in the image, return the resembling dog breed.
  • if neither is detected in the image, provide output that indicates an error.

You are welcome to write your own functions for detecting humans and dogs in images, but feel free to use the face_detector and human_detector functions developed above. You are required to use your CNN from Step 4 to predict dog breed.

Some sample output for our algorithm is provided below, but feel free to design your own user experience!

Sample Human Output

(IMPLEMENTATION) Write your Algorithm

In [106]:
### TODO: Write your algorithm.
### Feel free to use as many code cells as needed.

def run_app(img_path):
    ## handle cases for a human face, dog, and neither
    img = Image.open(img_path)
    
    if face_detector(img_path) > 0:
        preds = predict_breed_transfer(model_transfer, class_names, img_path)
        print("\n\nHello Human!")
        plt.imshow(img)
        plt.show()
        print('You look like a ',preds)
    elif dog_detector(img_path):
        preds = predict_breed_transfer(model_transfer, class_names, img_path)
        print("\n\nDog Detected!")
        plt.imshow(img)
        plt.show()              
        print("It looks like a ",preds)  
    else:
        plt.imshow(img)
        plt.show() 
        print("Error! Neither Human nor a Dog..")
In [107]:
for img_file in os.listdir(new_path):
    img_path = os.path.join(new_path, img_file)
    run_app(img_path)

Hello Human!
You look like a  American eskimo dog


Dog Detected!
It looks like a  Airedale terrier


Dog Detected!
It looks like a  Airedale terrier


Dog Detected!
It looks like a  American water spaniel
Error! Neither Human nor a Dog..


Dog Detected!
It looks like a  Basenji


Hello Human!
You look like a  Cane corso


Dog Detected!
It looks like a  Cavalier king charles spaniel


Dog Detected!
It looks like a  German wirehaired pointer


Dog Detected!
It looks like a  Greyhound

Step 6: Test Your Algorithm

In this section, you will take your new algorithm for a spin! What kind of dog does the algorithm think that you look like? If you have a dog, does it predict your dog's breed accurately? If you have a cat, does it mistakenly think that your cat is a dog?

(IMPLEMENTATION) Test Your Algorithm on Sample Images!

Test your algorithm at least six images on your computer. Feel free to use any images you like. Use at least two human and two dog images.

Question 6: Is the output better than you expected :) ? Or worse :( ? Provide at least three possible points of improvement for your algorithm.

Answer: (Three possible points for improvement)

  1. Trying more architectural changes with combination of more/less layers
  2. Hyperparameter tuning: learning_rate, epochs, batch_size
  3. Should have tried other image augmentation techniques for a better training
In [111]:
## TODO: Execute your algorithm from Step 6 on
## at least 6 images on your computer.
## Feel free to use as many code cells as needed.
test_path = "C:/Users/krishna/Desktop/Data Science/DeepLearning/Udacity/dogImages/final_test/"
## suggested code, below
for img_file in os.listdir(test_path):
    img_path = os.path.join(test_path, img_file)
    run_app(img_path)

Hello Human!
You look like a  Pharaoh hound


Hello Human!
You look like a  Xoloitzcuintli


Dog Detected!
It looks like a  Chihuahua


Dog Detected!
It looks like a  Belgian malinois


Hello Human!
You look like a  Poodle


Hello Human!
You look like a  Glen of imaal terrier


Dog Detected!
It looks like a  Labrador retriever


Dog Detected!
It looks like a  Japanese chin
In [ ]: